R 语言作图

赵鹏

(西交利物浦大学 健康与环境科学系)

1 课前预备

  1. 具备 R 语言基本常识
  2. 安装好 R 和 RStudio IDE: 英文指南， 中文指南
  3. 安装好扩展包：
install.packages(c('beginr', 'Epi', 'ggplot2','Rcmdr', 'ggplotgui', 'GrapheR', 'plotly', 'GGally', 'remotes', 'MSG', 'fun'))
remotes::install_github('pzhaonet/fecitr')
  1. 数据：准备自己的一个数据文件，以便课上练习。

2 课程目标

  1. 了解 R 语言在科研领域能绘制哪些图形
  2. 了解 R 语言作图的两种常见体系（即 base R 和 ggplot2 包）和两种实现手段（即图形界面和代码）
  3. 掌握常见统计图形的 R 语言方法

3 R 能做哪些图

3.1 从使用场景看

3.1.1 学术论文

3.1.2 演示文稿

3.1.3 其他效果

3.1.4 无处不在

3.2 从数据类型和统计方法看

4 两种实现手段

4.1 图形界面

library(GrapheR)
run.GrapheR()
library(Rcmdr)
ggplotgui::ggplot_shiny()
library(Deducer)
JGR()

4.2 代码

data(diet, package = "Epi")
library(ggplot2)
ggplot(diet) + 
  geom_point(aes(height, weight))

5 统计图形的构成

  • 图形

    • 绘图区
    • 坐标轴
    • 图例

  • 编号和标题

6 常见图形的做法

6.1 概率密度图

curve(dnorm(x), from = -5, to = 5)
segments(c(-1, 1), 0, c(-1, 1), dnorm(c(-1, 1)), 'red')

curve(dt(x, df = 10),  from = -5, to = 5)
abline(h = 0)
abline(v = qnorm(0.95), col = 'red')

curve(dchisq(x, df = 10),  from = 0, to = 30)
abline(h = 0)
abline(v = qchisq(0.95, df = 10), col = 'red')
Distribution p- q- d- r-
Binomial pbinom qbinom dbinom rbinom
Chi-Square pchisq qchisq dchisq rchisq
F pf qf df rf
Logistic plogis qlogis dlogis rlogis
Log Normal plnorm qlnorm dlnorm rlnorm
Normal pnorm qnorm dnorm rnorm
Poisson ppois qpois dpois rpois
Student t pt qt dt rt
Uniform punif qunif dunif runif
Weibull pweibull qweibull dweibull rweibull
Wilcoxon Rank Sum Statistic pwilcox qwilcox dwilcox rwilcox
Wilcoxon Signed Rank Statistic psignrank qsignrank dsignrank rsignrank

6.2 条形图，柱状图 (Bar chart)

tab1 <- table(diet$job)
barplot(tab1, horiz = TRUE, las = 1)

ggplot(diet) +
  geom_bar(aes(job)) +
  coord_flip()

tab2 <- table(diet[, c('job', 'chd')])
barplot(tab2, xlab = 'CHD', beside = TRUE, col = 1:3,
        legend.text = TRUE, 
        args.legend = list(
          x = 8, y = 100,
          legend = dimnames(tab2)$job,
          bty = 'n', col = 1:3))

ggplot(diet) +
  geom_bar(aes(chd, fill = job), position = 'dodge')


ggplot(diet) +
  geom_bar(aes(chd, fill = job)) +
  facet_grid( ~ energy.grp) # or facet_wrap( ~ energy.grp)

6.3 直方图，箱线图，小提琴图

hist(diet$weight, freq = FALSE)
lines(density(diet$weight))

ggplot(diet) +
  geom_histogram(aes(x = weight)) + 
  facet_grid(~job)

boxplot(diet$weight ~ diet$job)
diet_mean <- tapply(diet$weight, diet$job, mean, na.rm = TRUE)
points(1:3, diet_mean, pch = 4)

ggplot() + 
  geom_boxplot(aes(job, weight), data = diet) +
  geom_point(aes(1:3, diet_mean))

library(vioplot)
vioplot(diet$weight, horizontal = TRUE)
points(mean(diet$weight, na.rm = TRUE), 1, pch = 4, col = "red")

ggplot(diet, aes(x = weight, y = 0)) +
  geom_violin(fill = 'grey') +
  geom_boxplot(width = 0.1) +
  geom_point(aes(x = mean(diet$weight, na.rm = TRUE), y = 0), pch = 4)

6.4 散点图

ggplot(diet, aes(height, weight)) +
  geom_point(alpha = 0.1) +
  # geom_bin2d()
  geom_smooth(method = 'lm') +
  labs(x = 'Height(cm)', y = 'Weight (km)') +
  theme_bw()

7 更多图形

7.1 Base R

Function Description
barplot() Generate a bar chart.
boxplot() Generate a box plot (also called box-and-whisker plot).
contour() Generate a contour plot.
coplot() Generate a conditional plot.
curve() Generate a curve corresponding to a function.
dotchart() Generate a Cleveland dot plot.
hist() Generate a histogram.
image() Generate a grid of colored or gray-scale rectangles with colors corresponding to the values
matplot() Wrapper of plot() for plotting columns of one matrix against columns of another.
pairs() Generate a matrix of scatterplots.
persp() Generate a perspective plot of a surface
pie() Generate a pie chart.
plot() Generic plotting function, depending on the object’s class.
stripchart() Generate a one dimensional scatter plot.
stars() Generate a star plot or segment diagram of a multivariate data set.
symbols() Generate a plot with symbols (circles, squares, rectangles, stars, thermometers, and boxplots) representing data values.
ts.plot() Plot multiple Time Series.
abline() Add straight lines to a graph.
arrows() Add arrows to a graph.
axis() Add an axis to a graph.
box() Add a surrounding box to a graph.
legend() Add legends to a graph.
lines() Add connected line segments to a graph.
mtext() Add text to the margins of a graph.
points() Add points to a graph.
polygon() Add polygons to a graph.
rect() Add a rectangle to a graph.
rug() Add short lines representing data values to a graph.
segments() Add line segments to a graph.
text() Add text to a graph.
Common base R plotting functions

7.2 ggplot2

Function Description
geom_abline(), geom_hline(), geom_vline() Straight lines.
geom_area(), geom_ribbon() Area and ribbon plot.
geom_bar() Bar chart.
geom_histogram(), geom_freqpoly(), geom_bin2d(), geom_hex() 1D/2D histogram or frequency polygons.
geom_boxplot() Box plot.
geom_contour() Contour plot.
geom_crossbar(), geom_errorbar(), geom_errorbarh(), geom_linerange(),geom_pointrange() Data value with intervals.
geom_density(), geom_density2d() Distribution density curve/surface.
geom_line(), geom_path() Line chart.
geom_point() Scatterplot.
geom_polygon(), geom_rect(), geom_tile() Polygons/rectangles.
geom_rug() Rugs.
geom_segment(), geom_curve() Line segments and curves.
geom_smooth(), geom_quantile() Fitted lines.
geom_text() Text.
Common ggplot plotting functions
ggplot() +      # mandatory: data mapping
  geom_xxx() +  # mandatory: geometry shape
  coordinate_xxx() +  
  theme() +          
  facet_xxx() +      
  ...

7.3 两个例子

library(fecitr)
plot_summary(diet, if_box = TRUE)

library(GGally)
ggpairs(diet)

8 延伸阅读